GitHub & VS Code Setup
Before starting Module 0, Windows users with no prior experience should complete the environment setup. This covers Git installation, VS Code configuration, and cloning your first repository.
🛠 Open Setup Guide →Appendix A — Core Prompt Library
Copy and adapt these prompts. Always prefix with your context.md contents for best results.
PRD Generator
File: prompts/prd-draft.md · Use in: Module 5
You are an experienced product manager. Using the product context below, generate a one-page PRD for the following feature:
Feature: [FEATURE NAME]
Context: [PASTE context.md HERE]
The PRD must include:
1. Problem statement (2-3 sentences, user-centric)
2. Proposed solution (what, not how)
3. Success metrics (3 SMART KPIs)
4. Functional requirements as user stories (5-8 stories, As a... I want... So that...)
5. Out of scope (3-5 explicit exclusions)
6. Open questions (2-3 unresolved items)
Tone: clear, precise, no jargon. Length: under 500 words.User Story Splitter
File: prompts/story-split.md · Use in: Module 5
Split the following epic into independently deliverable user stories. Each story must:
- Follow the format: As a [user type], I want [action], so that [benefit]
- Be deliverable in a single sprint
- Have a clear acceptance criterion (one sentence)
- Be independent of the other stories where possible
Epic: [PASTE EPIC HERE]
Provide 5-8 stories. Flag any that have dependencies on each other.Interview Synthesis
File: prompts/interview-synthesis.md · Use in: Module 3
Synthesise the following user interview notes into a structured insight document.
Interview notes: [PASTE RAW NOTES]
Output format:
## Top Pain Points (ranked by frequency × severity)
1. [Pain point] — mentioned by [N] participants
2. ...
## Surprising Insights
- [Insight that challenges assumptions]
## Quotes Worth Sharing
- "[Quote]" — [role/segment]
## Opportunity Statement
How might we [verb] so that [user outcome]?
Tone: neutral, evidence-based. Length: under 300 words.RICE Scoring Assistant
File: prompts/rice-scoring.md · Use in: Module 6
Help me score the following features using the RICE framework.
Context: [PASTE context.md]
Features to score:
1. [Feature name + one-line description]
2. ...
For each feature, provide:
- Reach: estimated users affected per quarter
- Impact: 0.25 (minimal) / 0.5 (low) / 1 (medium) / 2 (high) / 3 (massive)
- Confidence: % based on evidence quality (100% = A/B tested, 50% = assumption)
- Effort: person-months
- RICE score: (Reach × Impact × Confidence) ÷ Effort
Output as a table. Add a one-sentence rationale for each Impact and Confidence score.Competitive Analysis Brief
File: prompts/competitive-brief.md · Use in: Module 4
Create a competitive analysis brief for the following competitors relative to [OUR PRODUCT].
Our product context: [PASTE context.md]
Competitors: [LIST 3-5 NAMES]
Focus area: [e.g., pricing, onboarding, specific feature]
For each competitor provide:
- Positioning statement (how they describe themselves)
- Key differentiators (2-3 bullets)
- Weaknesses or gaps (2-3 bullets)
- Pricing model (if known)
End with a gap analysis: what are we missing that competitors offer? What do we offer that they don't?
Output as a structured Markdown document.QA Loop Reviewer
File: prompts/qa-loop.md · Use in: All modules
Review the following [DOCUMENT TYPE: PRD / User Story / Research Synthesis / etc.] critically.
[PASTE DOCUMENT]
Identify:
1. Missing information that a reader would need
2. Assumptions that are stated as facts
3. Ambiguous language that could be misinterpreted
4. Internal contradictions
5. Anything that would be challenged in a stakeholder review
Rate overall quality: /10
Provide top 3 specific improvements.Retrospective Summariser
File: prompts/retro-summary.md · Use in: Module 9
Summarise the following sprint retrospective notes into an actionable brief.
Retro notes: [PASTE RAW NOTES]
Output:
## What went well (keep doing)
- [2-3 bullets]
## What didn't work (stop/change)
- [2-3 bullets]
## Action items
| Action | Owner | Due |
|--------|-------|-----|
| ... | ... | ... |
## One-sentence retrospective insight
[The most important learning from this sprint]
Length: under 250 words.Appendix B — Frameworks Cheat Sheet
RICE Prioritisation
- Reach: users/quarter
- Impact: 0.25 → 3 scale
- Confidence: % evidence quality
- Effort: person-months
- Score = (R × I × C) ÷ E
Hypothesis Format
- If [change we make]
- Then [observable outcome]
- Because [causal mechanism]
- Measured by [metric + threshold]
- Over [time period]
JTBD Format
- When [situation/trigger]
- I want to [motivation/goal]
- So I can [expected outcome]
- Focus: functional + emotional jobs
- Use for: discovery & positioning
OKR Format
- Objective: qualitative, inspiring direction
- KR: measurable, time-bound outcome
- 3-5 KRs per Objective
- KRs are outcomes, not activities
- "We will X as measured by Y"
MoSCoW Prioritisation
- Must: non-negotiable for launch
- Should: high value, workaround exists
- Could: nice-to-have, low effort
- Won't: out of scope for now
- Use for: MVP scoping
Now / Next / Later
- Now: in-progress commitments
- Next: committed next bets
- Later: directional, not committed
- No fake dates on Later items
- Review cadence: monthly
Kano Model
- Basic: expected; absence = dissatisfaction
- Performance: more = more satisfied
- Delight: unexpected; creates loyalty
- Indifferent: users don't care
- Use for: feature investment decisions
North Star Metric
- One metric that best captures value delivered
- Leading indicator of long-term retention
- Measurable, actionable, not a vanity metric
- Examples: Miro boards created, Slack messages sent
- Supporting metrics: 3-5 input metrics
Appendix C — Markdown Cheat Sheet
| Element | Markdown syntax | Renders as |
|---|---|---|
| Heading 1 | # Title | Large bold title |
| Heading 2 | ## Section | Section header |
| Heading 3 | ### Sub-section | Smaller header |
| Bold | **text** | text |
| Italic | *text* | text |
| Inline code | `code` | Monospace code |
| Code block | ```language ... ``` | Syntax-highlighted block |
| Bullet list | - item | Unordered list |
| Numbered list | 1. item | Ordered list |
| Link | [text](url) | Clickable link |
| Blockquote | > text | Indented quote block |
| Horizontal rule | --- | Dividing line |
| Table | | col | col | | Data table |
| Checkbox | - [ ] task | Unchecked checkbox |
| YAML front matter | --- key: value --- | Metadata block (top of file) |
Appendix D — Glossary
| Term | Definition |
|---|---|
| AI-native PM | A product manager who integrates AI into every stage of the PM workflow and builds AI-powered products by design, not as an afterthought |
| Context file | A structured Markdown document that gives an AI agent the product background it needs to produce accurate, relevant outputs |
| RAG | Retrieval-Augmented Generation — an AI architecture that retrieves relevant document chunks and injects them into a prompt before generation |
| QA loop | A second AI prompt that reviews and critiques the output of a first AI prompt before the PM acts on it |
| RICE | Reach × Impact × Confidence ÷ Effort — a quantitative prioritisation framework |
| JTBD | Jobs to Be Done — a framework for understanding user motivation: "When X, I want to Y, so I can Z" |
| North Star Metric | The single metric that best captures the value a product delivers to users; a leading indicator of long-term retention |
| PM OS | PM Operating System — the personal system of files, prompts, templates, and rituals a PM uses to work consistently and leverage AI effectively |
| Hallucination | When an AI model generates a confident but factually incorrect response |
| Prompt library | A curated collection of reusable, tested prompts organised by task type, stored as Markdown files |
| Decision log | A timestamped record of product decisions with rationale, alternatives considered, owner, and review date |
| MoSCoW | Must / Should / Could / Won't — a feature prioritisation and MVP scoping method |
| Kano model | A framework classifying features as Basic, Performance, Delight, or Indifferent based on their relationship to user satisfaction |
| OKR | Objectives and Key Results — a goal-setting framework where Objectives are qualitative direction and Key Results are measurable outcomes |
| Agentic AI | AI that can take multi-step actions autonomously (search, write, run code, call APIs) rather than responding to single queries |
| Chunking | The process of splitting documents into smaller units for storage and retrieval in a RAG system |
| Embedding | A numerical vector representation of text that captures semantic meaning, used for similarity search in RAG systems |
| PII | Personally Identifiable Information — data that can identify a specific individual; subject to data protection regulations |
| Funnel analysis | Measurement of user drop-off at each step of a defined user journey, used to identify conversion barriers |
| Opportunity statement | A "How Might We..." framing of a user problem that opens design space without prescribing solutions |
Appendix E — PM OS Starter Templates
context.md Template
---
product: [Product name]
owner: [Your name]
updated: YYYY-MM-DD
---
## Product Vision
[One sentence: what we're building and why it matters]
## Primary Users
[2-3 sentences: who they are, what they do, why they use this product]
## Top Problems We Solve Today
1. [Problem + evidence]
2. [Problem + evidence]
3. [Problem + evidence]
## Current Success Metrics
| Metric | Current | Target | Period |
|--------|---------|--------|--------|
| ... | ... | ... | ... |
## Key Constraints
- [Technical, compliance, resource, or strategic constraint]
- ...
## Decisions Already Made (off the table)
- [Decision + brief rationale]decisions.md Entry Template
## YYYY-MM-DD — [Decision title]
**Decision:** [What we decided, in one sentence]
**Rationale:** [Why this option, 2-3 sentences]
**Alternatives considered:**
- [Option A] — rejected because [reason]
- [Option B] — rejected because [reason]
**Owner:** [Name]
**Review date:** [YYYY-MM-DD or "no review needed"]
**Impact:** [What this decision affects]rituals.md Template
# PM Rituals
## Daily (10 min)
- Review context.md for current priorities
- Identify one task to use AI for today
- Log any decisions made yesterday
## Weekly (30 min — sprint prep)
- Update context.md with any changes
- AI-assisted backlog triage: score and rank top 10 items
- Capture new decisions in decisions.md
- Review last week's AI outputs — what was accurate, what wasn't
## Monthly (60 min — PM OS review)
- Prune prompt library: archive unused prompts
- Update templates based on recent learnings
- Review decision log: any decisions to revisit?
- Log a retrospective entry
## Quarterly (2 hrs — strategy refresh)
- Re-align context.md with new OKRs
- Review decision log for patterns
- Update competitive context
- Identify one new AI workflow to adopt this quarterAppendix F — Recommended Tools
| Category | Tool | Best for | PM OS role |
|---|---|---|---|
| AI chat + agents | Claude, ChatGPT, Gemini | Drafting, synthesis, QA loops | Primary AI interface |
| AI coding agents | Cursor, GitHub Copilot, Windsurf | Prototype building, data analysis | Technical execution |
| Markdown editors | Obsidian, Typora, VS Code | PM OS file management | Context & prompt storage |
| Product management | Linear, Jira, Productboard | Backlog, sprints, roadmaps | Execution tracking |
| Roadmapping | Airfocus, Productplan, Notion | Stakeholder-facing roadmaps | Now/Next/Later publishing |
| User research | Dovetail, Maze, UserTesting | Interview recording, analysis | Discovery synthesis input |
| Analytics | Mixpanel, Amplitude, PostHog | Funnel analysis, event tracking | Metrics and experiment data |
| Documentation | Notion, Confluence, GitBook | Knowledge base, specs, policies | RAG knowledge layer |
| Prototyping | Figma, v0.dev, Bolt | UI design, rapid prototypes | Discovery validation |
| Collaboration | Slack, Loom, Miro | Async communication, diagrams | Team alignment |
Tool selection principle: Choose tools that output or accept Markdown. This keeps your PM OS portable and AI-compatible regardless of which tools change over time. Never let a tool lock own your knowledge.